Presentation and Generation of Fault Diagnosis Knowledge for Mechano-Electronic Systems

Aim To focus on the concept description and fault diagnosis knowledge representation, acquisition and generation for complex mechano-electronic-electronic systems. Methods A concept-based structured language was proposed to describe the practical system. The concepts were classifted into three categories as entity concept, attribute con- cept and feature concept. Based on the entity concept, a concept dictionary and some concept trees were created. The inner structure of a complex mechano-electronic system was expressed through the concept tree. The fault information and fault diagnosis knowledge for the system were represented by the combining the three kinds of con- cepts mentioned above. Results Based on which a framework for FMS troubleshooting has been established. The framework encompasses the entire process from concept description, knowledge representation, to knowledge generation. An FMS fault diagnosis and maintenance system, in which a maintenance-database, an image-base are integrated with knowledge-bases, is established. Conclusion The meth- od is feasible and effective,and the fault diagnosis system is able to learn.